160 specialist AI agents for Claude Code, Cursor & Codex — one npm install puts them all into your IDE as MCP tools and parallel sub-agents. Local-first, BYOK, works offline with Ollama.
This repository also configures its own agents. See what timps-swarm tells them →
Agent Claude Code
Predict merge conflicts between two branches before merging, with file-level and line-level overlap analysis. Use the timpsmergeconflictpredictor MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Analyse a codebase for breaking-change APIs given a target library/framework upgrade. Returns a step-by-step migration checklist. Use the timpsmigrationpilot MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Scaffold React Native/Flutter/Expo apps with push notifications and deep links. Use the timpsmobileagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Create evaluation harnesses, adversarial test inputs, and RAGAS config for LLM models. Use the timpsmodelevaluator MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Monitor an ML model in production: drift, accuracy decay, latency, and cost — with alerting thresholds and a re-training trigger plan. Use the timpsmodelperfmonitor MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Configure Stripe/LemonSqueezy products, checkout flows, webhooks, and dunning emails. Use the timpsmonetizationagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate Prometheus metric definitions, Grafana dashboard JSON, alerting rules, and instrumentation code snippets. Use the timpsmonitoringagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
You are the n8n workflow agent sub-agent from the TIMPS Swarm (category: priority).
Agent Claude Code
Diagnose network problems: ping internet/Cloudflare, DNS resolution, WiFi signal, traceroute, open ports. Generates a fix script with DNS flush, WiFi reset, and DHCP renewal commands. Use the timpsnetworkmedic MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Analyse your observability bill (Datadog / New Relic / Grafana Cloud / Splunk) and recommend cardinality pruning, sampling, log-volume cuts, and tier migrations with projected savings. Use the timpsobservabilitycostoptimizer MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate a customised onboarding guide: most-touched files, architecture overview, key modules, recent PR walkthroughs. Use the timpsonboardingmentor MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Detect duplicate code blocks, god classes, magic numbers, and anti-patterns in a codebase. Use the timpspatterndetector MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate a realistic, organisation-tailored phishing simulation: emails, landing pages, payload tracking, and a click-through metrics plan. Strictly for authorised training use. Use the timpsphishingsimulator MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Read a failing CI/CD run and produce a minimal diff to make it green: cache tweaks, image bumps, secret injection, retry policies. Use the timpspipelinehealer MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Turn a podcast transcript / outline into SEO-friendly show notes: title, summary, chapter timestamps, key quotes, links, guest bio, and 3 tweet drafts. Use the timpspodcastshownoteswriter MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Write blameless postmortems: 5 Whys, root causes, P0-P2 action items, runbooks. Use the timpspostmortemagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Review a pull request diff for blockers, suggestions, and nitpicks. Returns a structured review with a 0-10 score and approve/reject verdict. Use the timpsprreviewer MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Privacy audit: count cookies per browser (Chrome, Firefox, Edge, Brave), list macOS app permissions (camera, mic, location, contacts). Returns a cleanup manifest — no data is deleted until user reviews. Use the timpsprivacycleaner MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Rewrite prompts with Chain-of-Thought, XML tags, and few-shot examples for better LLM output. Use the timpspromptengineer MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Scan prompts, system messages, and tool schemas for prompt-injection / jailbreak / data-exfiltration risk. Produces a severity-rated report and hardened prompt rewrites. Use the timpspromptinjectionscanner MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Identify quantum-vulnerable algorithms, generate Qiskit circuits, and produce NIST PQC migration plans. Use the timpsquantumready MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Design chunking/embedding/retrieval/reranking RAG pipelines for LlamaIndex or LangChain. Use the timpsragdesigner MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Run a RAGAS-style evaluation of a RAG pipeline: context precision/recall, faithfulness, answer relevance, MRR/NDCG, with a runnable eval harness and an improvement plan. Use the timpsragevaluator MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Build WebSocket/SSE/Redis Pub-Sub/Kafka server+client real-time systems. Use the timpsrealtimeagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: